The AI Revolution in Higher Education
The AI revolution in higher education is the compressed institutional crisis, visible from the November 2022 release of ChatGPT through the February 2026 Rethinking Clark Kerr lecture and beyond, in which large language models simultaneously disrupted every function of the American research university. Teaching: the lecture's informational purpose redundant. Research: the answering machinery accelerated, the question-formulation capacity untouched. Credentialing: employers shifting to portfolio-based assessment. Industry connection: frontier AI labs producing foundational research at corporate scale. General education: the broad intellectual formation most undervalued for fifty years suddenly the most economically necessary offering. No single function escaped; the combination made the crisis different in kind from every previous technological challenge the university had absorbed.
In the AI Story
The crisis arrived faster than the institution's governance structures were designed to process. Faculty senates that require semesters to approve a new course had to respond to a technology whose capabilities were doubling in months. Accreditation processes designed to ensure five-year stability had to evaluate institutions whose curricula were obsolete within a single academic year. The mismatch between the pace of AI capability and the pace of institutional response was not a minor operational challenge — it was the defining feature of the crisis.
Student behavior changed before faculty or administrators could respond. Students were using AI for coursework in 2023 in ways faculty had not yet imagined; by 2024, academic integrity policies were being rewritten in real time; by 2025, the question had shifted from how do we prevent AI use to how do we teach in an environment where AI use is assumed. The students were, as always, the constituency with the least institutional power and the most at stake, and they moved faster than the institutions they attended.
The hiring market moved in parallel. Major technology companies shifted toward skills-based hiring. Portfolio assessment became standard in creative and technical fields. The credential reckoning accelerated as employers learned to read what candidates had built — often with AI augmentation — rather than what their transcripts claimed they had studied. The enrollment declines already underway at non-selective four-year institutions deepened, and the hollowing of the middle of American higher education that demographers had projected for 2030 arrived in 2025.
By 2026, the framing of the crisis had converged on something close to the argument of this volume. The university's survival depended not on defending what it had been but on recovering what it was always meant to be — a place where human beings learn not what to think but how to decide. The judgment-cultivation function, neglected for fifty years because other functions generated the revenue, became the function that had to justify the institution's continued existence. Whether the institution could execute the turn at the pace the moment demanded remained, as this volume went to press, an open question.
Origin
The crisis's conventional starting point is November 30, 2022 — the public release of ChatGPT — but the underlying capabilities had been developing in research laboratories for years. The institutional reckoning gathered force through 2023–2024 as GPT-4, Claude, and Gemini demonstrated capabilities that could no longer be dismissed, and crystallized in early 2025 with the combination of Chan's Berkeley lecture, the Harvard tutor study, and enrollment data that confirmed the hollowing of the middle.
Key Ideas
Every function simultaneously. The first challenge in university history that struck teaching, research, credentialing, and industry connection at once.
Pace mismatch. Governance structures designed for five-year stability confronting a technology that transforms in months.
Students lead, institutions follow. The constituency with least institutional power moved fastest, forcing reactive rather than proactive responses.
Credential reckoning accelerates. Employer adoption of portfolio-based assessment hollowed the middle of American higher education.
Judgment cultivation as survival function. The framing that emerged by 2026 as the institutional response consolidated.
Appears in the Orange Pill Cycle
Further reading
- Gerald Chan, "Rethinking Clark Kerr" (Berkeley, February 2025)
- Ethan Mollick, Co-Intelligence: Living and Working with AI (Portfolio, 2024)
- Cathy N. Davidson, The New Education (Basic Books, 2017)
- Scott Galloway, Post Corona: From Crisis to Opportunity (Portfolio, 2020)